AdaptLearn

Inspiration

One-size-fits-all education is a relic. Learning is universal, but our methods shouldn't be limited by rigid formats. I was inspired to build AdaptLearn because I personally struggled with traditional lecture notes, slides, and videos. I found that existing platforms lacked the flexibility to adapt to how I actually process information. I realized that the "vibe" of a lesson matters: some need to hear it, some need to see it, and some need to build it. I wanted to unleash the full potential of every student by personalizing the medium, not just the content.

What it does

AdaptLearn is a multimodal study companion that "mutates" any material into the user’s perfect learning style: Visual: Converts dense text into knowledge maps and diagrams to visualize connections. Auditory: Transforms any learning materials into engaging audio forms that are concise and broken down for on-the-go listening. Reading/Writing: Condenses long-form content into high-retention summaries for those who learn best through text and note-taking. Kinesthetic: Leverages Gemini's code execution to generate real-time Python simulations so students can "play" with data, and transforms materials into various interactive mini-game forms.

How I built it

Native Multimodality: We utilized Gemini’s ability to process text, audio, and video in a single stream to ensure seamless transitions between learning modes. 1M Context Window: This allowed us to feed entire textbooks into the prompt, ensuring the mini-games and summaries were contextually accurate to the specific curriculum.

Challenges I ran into

The primary challenge was prompt precision, learning how to clearly articulate my exact needs to the AI to get the desired output. I also had to contend with API rate limits, which required me to be very strategic with my testing and forced me to refine my logic mentally before hitting the "run" button.

Accomplishments that I'm proud of

Although it is a rough prototype, I am incredibly proud that it can turn a single PDF into so many distinct forms. It proves that everyone can achieve the same result if given their preferred method. No one should be deemed "incapable" simply because the teaching format wasn't a match for their brain.

What I learned

I learned that Vibe Coding isn't just about throwing a list of wants at an AI. It’s about partnership, showing the model exactly what needs to be done and collaborating on the "how" to get there. It’s a shift from being a "coder" to being an architect of intent.

What's next for AdaptLearn

I plan to add more varied features under each learning method. While time constraints and rate limits slowed me down during this build, I am committed to keeping AdaptLearn an ongoing project to help more students find their perfect "learning vibe."

Built With

• gemini3